FLIGHT SAFETY FOUNDATION
JULY–AUGUST 1996 FLIGHT SAFETY DIGEST
SPECIAL DOUBLE ISSUE Dubrovnik-bound Flight Crew’s Improperly Flown Nonprecision Instrument Approach Results in Controlled-flight-into-terrain Accident
Report Special CFIT Safety FLIGHT SAFETY FOUNDATION For Everyone Concerned Flight Safety Digest With the Safety of Flight Vol. 15 No. 7/8 July–August 1996 Officers/Staff In This Issue Stuart Matthews Dubrovnik-bound Flight Crew’s Chairman, President and CEO Improperly Flown Nonprecision 1 Board of Governors Instrument Approach Results in James S. Waugh Treasurer Controlled-flight-into-terrain Accident Robert Reed Gray, Esq. An analysis of the terrain profile, accident General Counsel and Secretary aircraft configuration and flight profile Board of Governors determined that the accident aircraft’s ground- proximity warning system would not have ADMINISTRATIVE provided a warning to the crew, the official U.S. Nancy Richards Air Force report said. Executive Secretary FINANCIAL FSF CFIT Checklist 26 Brigette Adkins Accountant Worldwide Commercial Jet Transport 30 TECHNICAL Accident Rates Declined in 1995 Robert H. Vandel There were nine fatal accidents in 1995 Director of Technical Projects compared with 14 in 1994, and the fatal-accident Ellen Plaugher rate was lower than in all but two years since Secretary 1960.
MEMBERSHIP FAA Advisory Circular Outlines Ahlam Wahdan New Method for Designing Airport 34 Assistant to the Director of Membership and Development Pavement for Boeing 777 Book offers a new look at procedures for PUBLICATIONS avoiding midair collisions. Roger Rozelle Director of Publications MD-87 Crew Makes Mistaken Approach Girard Steichen To Military Airport 44 Assistant Director of Publications Rick Darby Sightseeing helicopter downed by fuel Senior Editor exhaustion at air show. C. Claire Smith Cover: Wreckage of aft fuselage and empennage at crash site of U.S. Air Editorial Consultant Force CT-43A (Boeing 737-200), Dubrovnik, Croatia, April 3, 1996. Photo: AP/Wide World Photos Karen K. Ehrlich Production Coordinator Flight Safety Foundation is an international membership organization Kathryn L. Ramage Librarian, dedicated to the continuous improvement of flight safety. Nonprofit Jerry Lederer Aviation Safety Library and independent, FSF was launched in 1945 in response to the aviation industry’s need for a neutral clearinghouse to disseminate objective safety information, and for a credible and knowledgeable body that would identify threats to safety, analyze the problems and recommend Jerome Lederer practical solutions to them. Since its beginning, the Foundation has President/Emeritus acted in the public interest to produce positive influence on aviation safety. Today, the Foundation provides leadership to more than 660 member organizations in 77 countries. Foreword
In 1991, Flight Safety Foundation (FSF) launched an international campaign to reduce the number of controlled-flight-into- terrain (CFIT) accidents, which was expanded later to include approach-and-landing accidents. An FSF-led task force, in counsel with the International Air Transport Association (IATA) and the International Civil Aviation Organization (ICAO), was formed and set as its goal to reduce by 50 percent the number of CFIT accidents by 1998. The FSF CFIT Task Force, which in 1994 received an Aviation Week & Space Technology Aerospace Laurel award for its work, includes representatives from airlines, equipment manufacturers, airframe manufacturers, professional aviation organizations and other technical and research organizations (see page iii).
CFIT accidents are the leading cause of commercial aviation fatalities. According to FSF CFIT Task Force statistics, CFIT and approach-and-landing accidents accounted for more than 80 percent of the fatalities in commercial transport accidents between 1979 and 1991. According to statistics compiled by McDonnell Douglas (see page 30), there were four commercial jet transport CFIT accidents in 1995. Approach-and-landing accidents in 1995 continued to be prominent among commercial jet transport accidents, with 20 such accidents worldwide.
As part of the overall CFIT-reduction plan, the FSF CFIT Task Force has developed several products including:
• An “FSF CFIT Safety Alert” distributed to thousands of operators worldwide at no cost to recipients that emphasized the importance of immediate and aggressive “pull-up” actions, unless circumstances (visual meteorological conditions and flight well-above flat terrain, for example) explicitly determined that a ground-proximity warning system (GPWS) warning was false. Early FSF CFIT Task Force findings had determined that flight crews in CFIT accidents often ignored GPWS warnings; delayed recommended pull-up procedures while trying to evaluate the accuracy of the GPWS warning; or failed to respond with sufficient aggressive pull-up action.
• An “FSF CFIT Checklist,” which helps pilots assess CFIT risks for specific flights and operations. To date, more than 30,000 copies of the FSF CFIT Checklist (reproduced following page 25) have been distributed by the Foundation worldwide with the help of sponsorships from FlightSafety International, Scandinavian Airlines System (SAS) and SimuFlite Training International at no cost to the recipients. The English version remains available at no charge. Additional sponsorships are being sought to distribute Chinese, Russian, French, Spanish and Arabic versions of the CFIT Checklist, which were translated by ICAO.
• A video training aid, CFIT: Awareness and Prevention, which examines several CFIT accidents and presents cockpit voice recorder (CVR) data and simulation to illustrate accident reduction strategies. More than 6,000 copies of the video were distributed worldwide in 1995 (4,500 at no cost, with major funding from Associated Aviation Underwriters (AAU), Gulfstream Aerospace and Jeppesen Sanderson). Copies of CFIT: Awareness and Prevention are available from the Foundation for US$30; and,
• The CFIT Education and Training Aid, a two-volume training package developed under the auspices of the FSF CFIT Task Force and produced by the Boeing Commercial Airplane Group. Scheduled for release in late 1996, the training aid examines a number of CFIT accidents and focuses on human factors and management issues. The comprehensive training aid, which contains data to design simulator escape-maneuver training, is intended to be a resource for developing policies, procedures and CFIT-avoidance standards. The training aid also includes the video Controlled Flight into Terrain: An Encounter Avoided, which analyzes a jet transport CFIT accident in detail and shows how such accidents can be avoided. A price for the CFIT Education and Training Aid has not been determined, but it is designed to be an affordable training product for a wide range of aviation operators.
FLIGHT SAFETY FOUNDATION • FLIGHT SAFETY DIGEST • JULY–AUGUST 1996 i The FSF CFIT Task Force has also made eight recommendations to ICAO. A recommendation to broaden requirements for the use of GPWSs has been adopted by ICAO, and the others are under review. The new GPWS standards, effective Dec. 31, 1998, require GPWS in all aircraft used in “international commercial and general aviation operations, where the MCTM [maximum certified takeoff mass] is in excess of 5,700 kilograms [12,500 pounds] ... or that are authorized to carry more than nine passengers.”
ICAO officials have indicated their support for the pending recommendations, which include:
• A call for color-shaded depictions of terrain altitude on instrument-approach charts;
• A warning against the use of three-pointer and drum-pointer altimeters;
• A recommendation that all countries adopt the use of hectopascals for altimeter settings;
• A call for the replacement of early-model GPWS equipment;
• A recommendation for the improved design and presentation of nonprecision instrument approach procedures with a standard three-degree approach slope, except where prohibited by obstacles;
• A call for the use of automated altitude call-outs; and,
• A recognition of the important CFIT-avoidance benefits provided by the global positioning system/global navigation satellite system (GPS/GNSS).
CFIT accidents involving commercial operators were examined in detail in a special April–May 1996 double issue of the Flight Safety Digest. The research focused on 156 CFIT accidents that occurred from 1988 through 1994. The report, “An Analysis of Controlled-flight-into-terrain (CFIT) Accidents of Commercial Operators, 1988 through 1994,” concluded that 75 percent of the 108 accident aircraft (for which data were available) were not equipped with a GPWS and that landing (descent)-phase and landing (approach)-phase accidents together accounted for nearly 70 percent of all CFIT accidents studied. The report also noted that nearly 60 percent of landing (approach)-phase accidents involved aircraft flying nonprecision approaches. Twenty- five percent of the approaches were very high frequency omnidirectional radio range (VOR) DME approaches.
Another FSF report, prepared for the Directorate-General of Civil Aviation of the Netherlands and published in the March 1996 Flight Safety Digest, determined that there was a “five-fold increase in accident risk among commercial aircraft flying nonprecision approaches compared to those flying precision approaches.” The report, “Airport Safety: A Study of Accidents and Available Approach-and-landing Aids,” said that the “chance for error by the crew is probably greater during a nonprecision approach compared to a precision approach, resulting from the increased workload and additional need to maintain situational awareness.” According to FSF CFIT Task Force statistics for the period 1988 through 1995, 15 CFIT accidents involved distance-measuring equipment (DME) stepdown approaches. The FSF CFIT Task Force concluded that such nonprecision approaches are unnecessarily hazardous.
The Foundation’s CFIT accident–reduction campaign, like other FSF-led efforts in the areas of wind shear accident avoidance and fatigue countermeasures, was designed to bring the resources of the industry together to work toward the common goal of improving an already admirable safety record.
— Stuart Matthews Chairman, President and CEO Flight Safety Foundation
ii FLIGHT SAFETY FOUNDATION • FLIGHT SAFETY DIGEST • JULY–AUGUST 1996 FSF CFIT Task Force
Following is a list of companies whose representatives launched the initial FSF CFIT Task Force. Since its formation, many additional companies and individuals have participated in the task force’s work.
Aeroformation Flight Safety Foundation Aerospatiale Inc. (now Aero International Regional) Flight Safety Foundation-CIS Air Canada FlightSafety International Air Line Pilots Association (ALPA) Gulfstream Aerospace Corp. Air Traffic Control Association (ATCA) International Air Transport Association (IATA) Airbus Industrie International Civil Aviation Organization (ICAO) Airclaims Group Ltd. International Federation of Air Line Pilots All Nippon Airways (ANA) Co. Ltd. Associations (IFALPA) Allegheny Airlines Inc. International Federation of Air Traffic Controllers’ Association (IFATCA) AlliedSignal Aerospace Japan Airlines American Express Bank Ltd. Jeppesen Sanderson Inc. Atlantic Southeast Airlines (ASA) KLM—Royal Dutch Airlines Australian Airlines The Kroger Co. Avions de Transport Regional Lockheed Martin Corp. BA Acrad Lufthansa German Airlines Beechcraft Middle East Airlines The Boeing Co. Monsanto Co. Britannia Airways Ltd. U.S. National Aeronautics and Space Administration British Airways (NASA) — Ames Research Center Canada 3000 Airlines Inc. National Business Aircraft Association (NBAA) Cessna Aircraft Co. Netherlands National Aerospace Laboratory (NRL) Comair Inc. Pakistan International Airlines Corp. Continental Airlines Inc. Raytheon Co. de Havilland Inc. SimuFlite Training International Delta Air Lines Inc. Smiths Industries Aerospace Dresser Industries Sundstrand Corp. Douglas Aircraft Co. United Airlines U.S. Federal Aviation Administration (FAA) Varig S.A. Finnair Oy Vereiningung Cockpit E.V.
FLIGHT SAFETY FOUNDATION • FLIGHT SAFETY DIGEST • JULY–AUGUST 1996 iii Additional CFIT-related Reading from FSF Publications
“Helicopter Strikes Water on Approach After Pilots Lose Altitude Awareness,” Helicopter Safety Volume 22 (July-August 1996). “Different Altimeter Displays and Crew Fatigue Likely Contributed to Canadian Controlled-flight-into-terrain Accident,” Accident Prevention Volume 52 (December 1995). “Commuter Crew’s Loss of Situational Awareness During Night Takeoff Results in Controlled Flight into Terrain,” Accident Prevention Volume 52 (October 1995). “Crew’s Failure to Monitor Terrain Clearance After Night Takeoff Results in Collision with Mountain,” Accident Prevention Volume 52 (September 1995). “Poorly Flown Approach in Fog Results in Collision with Terrain Short of Runway,” Accident Prevention Volume 52 (August 1995). “Failure to Intercept Final Approach Course, Improperly Performed IFR Approach Cited in Fatal Collision with Terrain,” Helicopter Safety Volume 21 (May-June 1995). “Aircraft Descended Below Minimum Sector Altitude and Crew Failed to Respond to GPWS as Chartered Boeing 707 Flew into Mountain in Azores,” Accident Prevention Volume 52 (February 1995). “Breakdown in Coordination by Commuter Crew During Unstabalized Approach Results in Controlled-flight- into-terrain Accident,” Accident Prevention Volume 51 (September 1994). “Lack of Management Oversight Cited in Controlled-flight-into-terrain Accident of FAA Aircraft,” Accident Prevention Volume 51 (August 1994). “Captain Stops First Officer’s Go-around, DC-9 Becomes Controlled-flight-into-terrain (CFIT) Accident,” Accident Prevention Volume 51 (February 1994). “Cockpit Coordination, Training Issues Pivotal in Fatal Approach-to-Landing Accident,” Accident Prevention Volume 51 (January 1994). “Fatal Commuter Crash Blamed on Visual Illusion, Lack of Cockpit Coordination,” Accident Prevention Volume 50 (November 1993). “Anatomy of a Mountain Crash: Error Chain Leads to Tragedy,” Accident Prevention Volume 49 (October 1992).
iv FLIGHT SAFETY FOUNDATION • FLIGHT SAFETY DIGEST • JULY–AUGUST 1996 Dubrovnik-bound Flight Crew’s Improperly Flown Nonprecision Instrument Approach Results in Controlled-flight-into-terrain Accident
An analysis of the terrain profile, accident aircraft configuration and flight profile determined that the accident aircraft’s ground-proximity warning system would not have provided a warning to the crew, the official U.S. Air Force report said.
FSF Editorial Staff
On April 3, 1996, the crew of the U.S. Air Force CT-43A The accident flight crew and aircraft were assigned to the Air (Boeing 737-200) was flying a nondirectional radio beacon Force 76th Airlift Squadron (76 AS), 86th Operations Group (NDB) approach in instrument meteorological conditions (86 OG), 86th Airlift Wing (86 AW) at Ramstein Air Base (IMC) to Runway 12 at the Cilipi Airport, Dubrovnik, (AB), Germany. The crew’s mission was to transport U.S. Croatia, when the aircraft collided with a 701-meter (2,300- Department of Commerce Secretary Ronald H. Brown and a foot) mountain. All six crew members and all 29 passengers delegation of U.S. industry executives from Zagreb, Croatia, were killed in the accident. to various locations in Bosnia-Herzegovina and Croatia during a three-day period. The accident occurred on the first day, on As the aircraft crossed the final approach fix (FAF), it had an the fourth leg of a five-leg trip. airspeed of approximately 209 knots (387 kilometers per hour), which was 30 knots (56 kilometers per hour) faster than the The itinerary was changed four times before the flight began. airspeed recommended in the airplane’s flight manual for The crew planned the mission on April 1 (two days before the crossing the FAF. accident flight), based on the information in Change 1 to their itinerary. A stop in Dubrovnik was not listed in the change. As they flew the approach, the crew did not track the final- approach course of 119 degrees, but instead tracked a course The pilot of the flight “was known for very thorough mission of 110 degrees. They maintained this course until the aircraft planning and briefing,” the report said. “Mission planning collided with the mountain, 1.8 nautical miles (NM) (3.4 typically included flight plan preparation and review of kilometers) north of the threshold of Runway 12. approaches in accordance with squadron policy; however, the squadron mission briefing guide did not specifically include A U.S. Air Force Accident Investigation Board concluded in approach review. The crew would not have flight-planned for its final accident report that the accident was caused by “a the Dubrovnik leg, because Dubrovnik was not part of the failure of command, air crew error and an improperly designed Change 1 mission,” the report said. approach procedure.” At 1945 hours local time that same day, Change 2 to the The accident aircraft was not equipped with either a cockpit itinerary was transmitted to the 76 AS. This change added voice recorder (CVR) or a flight data recorder (FDR), the report Dubrovnik as a stop on the first day’s itinerary. “It could not said. be confirmed if the [accident] pilots received the information
FLIGHT SAFETY FOUNDATION • FLIGHT SAFETY DIGEST • JULY–AUGUST 1996 1 Photo: AP/Wide World Photos Wreckage of the U.S. Air Force CT-43A that crashed near Dubrovnik, Croatia, on April 3, 1996, killing U.S. Department of Commerce Secretary Ron Brown and all 34 others on board.
concerning Change 2 that night,” the report said. “The approximately 2200, the copilot called the operations center [accident] copilot’s spouse testified that the copilot received a at Ramstein AB and asked for the latest mission change. “He phone call that evening from the [accident] pilot who requested was verbally briefed on Change 4 and was faxed a copy of that the copilot report to the squadron earlier than planned the Change 4; however, he received only the cover sheet,” the next morning, possibly due to a mission change.” report said.
On the morning of April 2 (the day before the accident flight), Change 4 resulted in the following itinerary for the first day’s the crew prepared for their mission at the squadron, before mission: Zagreb to Tuzla, Bosnia; Tuzla to Split, Croatia; Split departing on a pre-positioning flight to Zagreb. During their to Tuzla; Tuzla to Dubrovnik; and Dubrovnik to Zagreb (Figure planning, the copilot requested a change (Change 3) to the 1, page 3). This change added Split to the itinerary, which was second day’s itinerary, which was approved. [That change required because there was not enough ramp space at Tuzla would allow the crew to return to Zagreb while the commerce for the accident aircraft to stay on the ground for the scheduled secretary’s group was in Sarajevo.] “The copilot also asked a time of seven hours and 20 minutes. Thus, the crew would squadron flight planner to build flight plans for the Dubrovnik drop off their passengers in Tuzla, fly to Split and return later stop,” the report said. “This indicates the pilots were aware of to pick up their passengers in Tuzla before continuing. Change 2 before they departed for Zagreb. They also may have had the opportunity to secure mission planning information While the crew rested in Zagreb, their aircraft was en route for Dubrovnik before they departed Ramstein AB on April 2, from Cairo, Egypt. The pilot of the Cairo flight had received 1996,” the report said. information on the latest change to the accident crew’s mission, the report said. “The Cairo pilot performed some mission The crew departed on another aircraft (flown by another flight preparation for the next day’s missions during this flight, crew) from Ramstein AB at 1232 and arrived in Zagreb at because he knew the [accident] crew was already in crew rest,” 1400. They arrived at approximately 1500 at the hotel where the report said. “He [the Cairo pilot] did not know whether they would remain overnight. Their planned departure time they had already received Change 4 and believed he could was between 0330 and 0400 on the morning of April 3. At help with the planning.”
2 FLIGHT SAFETY FOUNDATION • FLIGHT SAFETY DIGEST • JULY–AUGUST 1996 Itinerary for U.S. Air Force CT-43A, April 3, 1996
SLOVENIA HUNGARY Arad Ljubljana Pecs ROMANIA Zagreb
1 Timisoara CROATIA Rijeka Novi Sad BOSNIA- HERZEGOVINA Tuzla Banja Luka 3 Belgrade 5 SERBIA 2 Sarajevo Numbers refer to the legs scheduled to be flown on Split 4 April 3. No. 4 was the Adriatic Sea accident flight; No. 5 did not take place because of Ploce the accident.
MONTENEGRO Dubrovnik ITALY Titograd
Source: U.S. Air Force Figure 1
The report continued: “The Cairo pilot prepared flight plan (4,800 meters [approximately three miles] visibility) for the information that included proposed routing for three of the approach to Dubrovnik.” five legs (including Tuzla to Dubrovnik). He also provided communications information … and removed the Dubrovnik The accident copilot told the Cairo pilot that “he had received approach procedure, published by Jeppesen Sanderson Inc., a mission change by fax to the hotel earlier, but he had received from the aircraft’s publications kit. He planned to deliver these only the fax cover sheet, with no attached information. They to the [accident] crew when he met with them in Zagreb.” talked for about five minutes,” the report said.
The accident aircraft arrived in Zagreb at 2320. The Cairo pilot On April 3 (the day of the accident flight), the accident crew went to the hotel where the accident crew was staying, arriving assembled in the hotel at approximately 0330, so that they there between approximately 0030 and 0100 [April 3]. “He could report to the aircraft at 0400. The crew contacted the [the Cairo pilot] called the [accident] copilot’s room to arrange operations center at Ramstein AB and was verbally given for the delivery of the copilot’s personal clothing items [which Change 4. “A fax containing this change was retransmitted to were stowed on board the aircraft from Cairo], as well as the hotel,” the report said. the mission change information, planning data and the Dubrovnik instrument-approach procedure,” the report said. At 0403, a weather briefing was faxed from the Ramstein AB weather station to the hotel. “The weather briefing included The report noted: “The Cairo pilot testified that he believed Tuzla, Dubrovnik and Zagreb,” the report said. “The forecast the [accident] copilot was asleep when he called. After he weather for Dubrovnik was: wind 140 degrees at 10 knots called, the Cairo pilot delivered the information to the copilot’s [18.5 kilometers per hour], unrestricted visibility, ceiling room. The Cairo pilot said the copilot was dressed in his jeans broken at 2,500 feet [762 meters] and broken at 8,000 feet [1.5 but looked as if he had just awoke. The Cairo pilot gave the statute miles (SM)/2.4 kilometers], with temporary conditions copilot his clothing and the paperwork he had prepared, of rain. It is not known if the crew received this weather especially noting the unusually high weather requirements briefing prior to departure from the hotel,” the report said.
FLIGHT SAFETY FOUNDATION • FLIGHT SAFETY DIGEST • JULY–AUGUST 1996 3 The 51-minute flight to Tuzla was uneventful, but “while flying the approach at Tuzla, the approach controller notified the crew that they were well left of the final-approach course,” the report said. “The crew responded that they were correcting. After joining the final-approach course, the crew requested a 360- degree turn in order to ‘lose a couple thousand feet,’” the report said.
The crew landed at Tuzla at 0715. The passengers deplaned, and the aircraft departed 37 minutes later, at 0752. The departure was 32 minutes behind schedule.
At 0832, the aircraft arrived at Split and the crew had the aircraft refueled. “They departed Split at 1156, four minutes ahead of schedule; however, the scheduled 40-minute return flight to Tuzla actually took 51 minutes,” the report said. “The extra time was required because the crew flight-planned into Bosnia-Herzegovina through a closed corridor.” The corridor that the crew was required to use added 90 NM (167 kilometers) to the flight.
At 1247, the aircraft landed at Tuzla, where the passengers reboarded. “The [Department of Commerce] added two Croatian nationals to the party, bringing the total passengers to 29,” the report said.
The accident flight departed Tuzla for Dubrovnik at 1355. “After takeoff, the [accident] crew checked in with Tuzla Departure Control, who cleared them to flight level (FL) 160 (16,000 feet [4,880 meters]), following the filed standard instrument departure,” the report said. “The crew also asked Tuzla departure for approval to turn left to avoid thunderstorms.”
Boeing 737-100/200 The report continued: “At 1402, radar service for IFO21 was terminated by Tuzla Departure Control, and IFO21 was The short-/medium-haul Boeing 737 first entered service in 1967. The 737-100 seated 115, which was increased instructed to contact Zagreb NATO [North Atlantic Treaty to 130 with the introduction of the 737-200. The twin- Organization] Air Traffic Control (ATC) and monitor MAGIC turbofan 737 (U.S. Air Force designation T-43A), has a (the call-sign for a NATO E-3 Airborne Early Warning [AEW] maximum takeoff weight of 52,390 kilograms (115,500 aircraft). They checked in with Zagreb NATO ATC at 1406 pounds) and a maximum cruising speed of 462 knots and were cleared to FL 190 [19,000 feet (5,795 meters)], then (856 kilometers per hour). It has a range of 1,855 nautical to FL 250 [25,000 feet (7,625 meters)].” miles (2,136 miles/3,437 kilometers) with 115 passengers cruising at 33,000 feet (10,060 meters). IFO21 also established contact with the NATO E-3 AEW Source: Jane’s All the World’s Aircraft aircraft, the report said. Good radio communications and radar contact were established with the AEW aircraft (which was responsible for flight monitoring and threat warning in After the crew arrived at their aircraft, they contacted the Croatia and Bosnia-Herzegovina airspace). “The U.S. Air Force Europe (USAFE) Meteorological Service responsibilities of NATO E-3 AEW aircraft do not include (METRO) via high-frequency radio at 0440, the report said. course correction when an aircraft is flying an approach to They received a briefing on the weather for Tuzla (the first an airfield,” the report said. leg of their itinerary) and Zagreb (their alternate for the first leg). At 1404, IFO21 contacted the operations center at Ramstein AB. “[The accident crew] reported their takeoff time from Operating with a call sign of “IFO21,” the accident aircraft Tuzla and asked if there were any mission changes beyond departed Zagreb at 0624. “This was 24 minutes later than the Change 4,” the report said. “They were informed there were planned departure, because the [Department of] Commerce no additional mission changes and were cautioned about the party arrived late,” the report said. possibility of fog at the Dubrovnik airport.” The crew
4 FLIGHT SAFETY FOUNDATION • FLIGHT SAFETY DIGEST • JULY–AUGUST 1996 acknowledged the weather information and terminated the IFO21 contacted Dubrovnik Approach/Tower at 1446 and was contact. cleared direct to the Kolocep (KLP) NDB [15.9 kilometers (9.9 miles) from the Cavtat (CV) NDB, 11.8 NM from runway The accident crew then contacted USAFE METRO and threshold]. After opposite-direction traffic had been cleared, requested the forecast at Dubrovnik for their planned arrival IFO21 was cleared to descend to 5,000 feet (1,525 meters). time of 1440. “USAFE METRO provided them with the The accident flight began a descent from 10,000 feet following weather information: wind 110 degrees at 12 knots approximately 16 NM (30 kilometers) from KLP. [22 kilometers per hour], ceiling at 500 feet [152 meters] broken, 2,000 feet [610 meters] overcast, 8,000 feet [2,440 meters] “Calculations from groundspeed data indicate the pilot flew overcast, altimeter 29.85 inches of mercury, pressure altitude approximately 224 [knots]–243 knots [414 kilometers per +609 feet [186 meters], temperature +[52 degrees F (11 degrees hour–450 kilometers per hour] indicated airspeed during this C)], [approximately five SM (8,000 meters)] visibility and rain,” descent,” the report said. “Although there are no flight manual the report said. “IFO21 made a pilot weather report to the performance charts that depict this specific descent speed, the USAFE METRO, reporting overall cloudy conditions with little charts for 250 knots [463 kilometers per hour] indicate this indication of thunderstorm activity.” descent would take approximately 16 NM with idle power and landing gear up.” As the aircraft flew through Bosnia-Herzegovina airspace en route to Croatia airspace, the AEW aircraft called the crew and The report noted: “Engine anti-ice was ‘on’ at impact. When advised that they were going out of the approved corridor. Under engine anti-ice is on, the operating manual requires an increased Bosnia-Herzegovina flight regulations, there were only three power setting. Use of speed brakes would offset the effect of corridors that aircraft were permitted to use, and one of those the increased power setting. Landing gear down would have — the BEAR corridor — was open only at increased the rate of descent; however, the specific times. The BEAR corridor, which tracking data shows the aircraft descended the flight crew had selected as part of their “The crew’s original in approximately the distance and time that route, was closed at the time of IFO21’s would indicate the landing gear was up. The flight. “The crew’s original flight plan and flight plan and their crew leveled the aircraft for about one their request for the [closed] corridor indicate minute, where the speed slowly decelerated, they were unfamiliar with the corridor times request for the [closed] possibly indicating the power remained of operation,” the report said. corridor indicate they reduced in an attempt to slow the aircraft for [approach] configuration.” IFO21 asked Zagreb NATO ATC to provide were unfamiliar with a radar vector to the approved corridor. At 1452, the crew told Dubrovnik After receiving an initial vector, IFO21 the corridor times Approach/Tower that they were 16 NM resumed navigation along the approved of operation.” from the airport. “They were cleared to corridor. “This unplanned routing added descend to 4,000 feet [1,220 meters] and approximately 15 [minutes]–16 minutes to told to report crossing the KLP beacon,” IFO21’s flight time,” the report said. the report said. “They began the descent and descended through 4,100 feet [1,250 meters] as they crossed KLP. They As the accident crew approached the Bosnia-Herzegovina were still [flying] too fast to fully configure with landing border, “Zagreb NATO ATC transferred control of IFO21 to flaps before beginning the final approach, as required by Air Croatian Zagreb Center civilian controllers (Zagreb Center),” Force directives,” the report said. the report said. Then IFO21 was cleared direct to the Split very high frequency omnidirectional radio range (VOR). At 1450, the pilot of a Croatian aircraft on the ground at “Zagreb Center cleared the [accident flight] to descend to Dubrovnik radioed the accident crew and asked the crew to [FL] 210 (21,000 feet [6,405 meters]), to be level 25 NM contact him on a frequency of 123.47 megahertz (MHz). (This [46 kilometers] before Split,” the report said. was a different frequency from the one being used for the approved tower communication.) The accident aircraft was equipped IFO21 crossed Split at 1434 and was cleared to descend from with two very high frequency (VHF) communication transceivers; FL 210 to FL 140 (14,000 feet [4,270 meters]). “[The accident therefore the crew could communicate with the pilot on the crew] started a normal descent and received further descent ground and simultaneously monitor the tower frequency. clearance from FL 140 to FL 100 [10,000 feet (3,050 meters)] at 1439,” the report said. “After the aircraft reached FL 100 at The pilot on the ground at Dubrovnik had landed one hour 1445, south of Split VOR, Zagreb Center transferred control earlier with the U.S. Ambassador to Croatia and the Prime to Dubrovnik Approach/Tower, a nonradar approach facility. Minister of Croatia, who were awaiting the arrival of Secretary The Dubrovnik Approach/Tower provides air traffic control Brown. The pilot later testified that he told the accident crew services based on vertical, lateral or longitudinal separation, that he had landed one hour earlier, and the weather was at the rather than radar monitoring,” the report said. minimum required for the approach. “He [the pilot on the
FLIGHT SAFETY FOUNDATION • FLIGHT SAFETY DIGEST • JULY–AUGUST 1996 5 ground] also testified that he told the IFO21 pilot that if he minimums as the NDB approach procedure. Minimums for the had to execute a missed approach, he should proceed to Split,” NDB Runway 12 approach were: minimum descent altitude the report said. “The [Dubrovnik] pilot testified [that] he 2,150 feet (656 meters) and visibility three SM (4.8 kilometers).] conversed with the IFO21 pilot for not more than 20 seconds. [He] also testified that IFO21 acknowledged the conversation.” Investigators could not determine whether the accident pilot or the flight mechanic answered the Dubrovnik pilot’s call The report added: “Other testimony indicated that [the concerning the special circling procedure. “Testimony from Dubrovnik pilot] contacted IFO21 two times and told the the [Dubrovnik] pilot indicated that it was not the same voice [accident] crew about a circling procedure that [he] had used to that was making radio calls on the tower frequency,” the report land. This procedure was not published in the Jeppesen approach said. “Other testimony indicated that the mishap copilot was procedure.” [Although a special circling procedure was not talking on the tower frequency.” depicted on the Dubrovnik NDB Runway 12 approach chart, the Jeppesen chart package for Dubrovnik included a special At the time of the accident, the NDB approach to Runway 12 circling procedure to Runway 30. This procedure had the same (Figure 2) was the only instrument-approach procedure
Illustration of NDB Approach to Runway 12, Cilipi Airport, Dubrovnik, Croatia
318 KLP 10 NM _._ ._.. ._ _. (Morse-code identification)
119 °
397 CV _._. ..._ City of Dubrovnik (Morse-code identification)
119 299 ° ° Approximate accident site 1.8 NM from airport INFORMATION NOT TO SCALE.
Abbreviations: NDB — Nondirectional radio beacon FAF — Final approach fix Cilipi Airport MDA — Minimum descent altitude MAP — Missed-approach point NM — Nautical miles
KLP NDB 4,000 feet FAF CV NDB/Compass locator 119 ° NOT FOR5.2 %NAVIGATION gradient MAP/2,149 feet
MDA/2,150 feet RUNWAY 12 518 feet 11.8 NM 1.9 NM Distance to displaced threshold
Source: Adapted from Accident Investigation Board Report United States Air Force CT-43A 3 April 1996 Dubrovnik, Croatia. This FSF-produced illustration is provided for educational purposes only. This illustration is not approved or intended for navigational purposes by any government or nongovernment entity, including but not limited to civil aviation authorities and commercial companies.
Figure 2
6 FLIGHT SAFETY FOUNDATION • FLIGHT SAFETY DIGEST • JULY–AUGUST 1996 Reconstruction of Final Approach Flown by Accident Crew, U.S. Air Force CT-43A, Dubrovnik, Croatia, April 3, 1996 10
Data labels are altitudes in 5200 feet mean sea level (MSL) from Mode C transponder
4700
3900 5 3200 2600 Kolocep (KLP) nondirectional beacon [NDB] 2300 Initial impact point (final approach fix [FAF]) 2200 at 2,172 feet (662 2200 meters) MSL (Minimum descent Cavtat (CV) [NDB] [compass] locator altitude [MDA] = [missed-approach point (MAP)] 2,150 feet [656 meters] MSL)
0 Runway 12/30